[FFmpeg-devel] [PATCH] swscale: [loongarch] Fix checkasm-sw_yuv2rgb failure.

--- libswscale/loongarch/swscale_init_loongarch.c | 104 +- libswscale/loongarch/yuv2rgb_lasx.c | 4 +- 2 files changed, 56 insertions(+), 52 deletions(-) diff --git a/libswscale/loongarch/swscale_init_loongarch.c b/libswscale/loongarch/swscale_init_loongarch.c index

Re: [FFmpeg-devel] [PATCH] swscale: [loongarch] Fix undeclared functions prob.

Ping. > 2024年5月15日 16:57，金波 写道： > > Look good to me. > > 2024-05-08 18:07:49 "yinshiyou-hf" 写道： >> Compile with '--disable-lasx', ‘lumRangeFromJpeg_lasx’ undeclared. >> --- >> libswscale/loongarch/swscale_init_loongarch.c | 2 ++ >> 1 file changed, 2 insertions(+) >> >> diff --git a/libswscale

[FFmpeg-devel] [PATCH] swscale: [loongarch] Fix undeclared functions prob.

Compile with '--disable-lasx', ‘lumRangeFromJpeg_lasx’ undeclared. --- libswscale/loongarch/swscale_init_loongarch.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/libswscale/loongarch/swscale_init_loongarch.c b/libswscale/loongarch/swscale_init_loongarch.c index 3a5a7ee856..4af62ad9f8 100

Re: [FFmpeg-devel] Add optimization in swscale for LA.

> 2024年3月27日 03:31，Michael Niedermayer 写道： > > On Tue, Mar 26, 2024 at 11:11:00AM +0800, Shiyou Yin wrote: >> >>> 2024年3月16日 11:03，Shiyou Yin 写道： >>> >>> [PATCH 1/3] swscale: [LA] Optimize range convert for yuvj420p. >>> [PATCH 2/3] sw

Re: [FFmpeg-devel] Add optimization in swscale for LA.

> 2024年3月16日 11:03，Shiyou Yin 写道： > > [PATCH 1/3] swscale: [LA] Optimize range convert for yuvj420p. > [PATCH 2/3] swscale: [LA] Optimize yuv2plane1_8_c. > [PATCH 3/3] swscale: [LA] Optimize swscale funcs in input.c > > ___ &

[FFmpeg-devel] [PATCH 3/3] swscale: [LA] Optimize swscale funcs in input.c

+} diff --git a/libswscale/loongarch/input_lsx.c b/libswscale/loongarch/input_lsx.c new file mode 100644 index 00..1bb04457bb --- /dev/null +++ b/libswscale/loongarch/input_lsx.c @@ -0,0 +1,65 @@ +/* + * Copyright (C) 2024 Loongson Technology Corporation Limited + * Contributed by Shiyou Yin

[FFmpeg-devel] [PATCH 2/3] swscale: [LA] Optimize yuv2plane1_8_c.

--- libswscale/loongarch/output.S | 254 +- libswscale/loongarch/output_lasx.c| 23 +- libswscale/loongarch/output_lsx.c | 22 +- libswscale/loongarch/swscale_init_loongarch.c | 12 +- libswscale/loongarch/swscale_loongarch.h | 29 +-

[FFmpeg-devel] Add optimization in swscale for LA.

[PATCH 1/3] swscale: [LA] Optimize range convert for yuvj420p. [PATCH 2/3] swscale: [LA] Optimize yuv2plane1_8_c. [PATCH 3/3] swscale: [LA] Optimize swscale funcs in input.c ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman

[FFmpeg-devel] [PATCH 1/3] swscale: [LA] Optimize range convert for yuvj420p.

--- libswscale/loongarch/swscale.S| 368 ++ libswscale/loongarch/swscale_init_loongarch.c | 33 ++ libswscale/loongarch/swscale_loongarch.h | 11 + libswscale/swscale_internal.h | 1 + libswscale/utils.c| 6 +-

Re: [FFmpeg-devel] [PATCH v3 7/7] avutil/la: Add function performance testing

> 2023年5月25日 10:36，Hao Chen 写道： > > > 在 2023/5/24 下午7:03, Rémi Denis-Courmont 写道: >> >> Le 24 mai 2023 10:39:59 GMT+03:00, Hao Chen a écrit : >>> 在 2023/5/20 下午5:38, Rémi Denis-Courmont 写道: Le lauantaina 20. toukokuuta 2023, 10.27.19 EEST Hao Chen a écrit : > From: yuanhecai >

Re: [FFmpeg-devel] [PATCH v4 1/7] avcodec/la: add LSX optimization for h264 idct.

> 2023年5月25日 05:28，Michael Niedermayer 写道： > > On Wed, May 24, 2023 at 03:48:27PM +0800, Hao Chen wrote: >> From: Shiyou Yin >> >> loongson_asm.S is LoongArch asm optimization helper. >> Add functions: >> ff_h264_idct_add_8_lsx >> ff_h264

Re: [FFmpeg-devel] Add LSX optimization in avcodec and swscale.

> 2023年5月20日 15:27，Hao Chen 写道： > > Retrigger the fate test. > v1: Add LSX optimization in avcodec and swscale, due to the 2K series CPUs > only support lsx. > v2: Modified the implementation of some functions and added support for the > checkasm --bench feature. > v3: Fix whitespace errors i

Re: [FFmpeg-devel] Add LSX optimization in avcodec and swscale.

> 2023年5月17日 15:03，Hao Chen 写道： > > v1: Add LSX optimization in avcodec and swscale, due to the 2K series CPUs > only support lsx. > v2: Modified the implementation of some functions and added support for the > checkasm --bench feature. > > [PATCH v2 1/7] avcodec/la: add LSX optimization for

Re: [FFmpeg-devel] [PATCH v2 1/7] avcodec/la: add LSX optimization for h264 idct.

> 2023年5月17日 15:03，Hao Chen 写道： > > From: Shiyou Yin > > loongson_asm.S is LoongArch asm optimization helper. > Add functions: > ff_h264_idct_add_8_lsx > ff_h264_idct8_add_8_lsx > ff_h264_idct_dc_add_8_lsx > ff_h264_idct8_dc_add_8_l

Re: [FFmpeg-devel] [PATCH v1 3/6] avcodec/la: Add LSX optimization for h264 chroma and intrapred.

rch.h > @@ -0,0 +1,43 @@ > +/* > + * Copyright (c) 2023 Loongson Technology Corporation Limited > + * Contributed by Shiyou Yin <mailto:yinshiyou...@loongson.cn>> > + * > + * This file is part of FFmpeg. > + * > + * FFmpeg is free software; you can redistribute it and/or > + * m

Re: [FFmpeg-devel] [PATCH] avutil: [LA] use getauxval to do runtime check.

> 2023年2月23日 15:48，Steven Liu 写道： > > Shiyou Yin 于2023年2月14日周二 20:26写道： >> >> Replace cpucfg with getauxval to avoid crash in case of >> some processor capabilities are not supportted by kernel used. >> --- >> libavutil/loongarch/cpu.c | 24 ++

[FFmpeg-devel] [PATCH] avutil: [LA] use getauxval to do runtime check.

Replace cpucfg with getauxval to avoid crash in case of some processor capabilities are not supportted by kernel used. --- libavutil/loongarch/cpu.c | 24 1 file changed, 8 insertions(+), 16 deletions(-) diff --git a/libavutil/loongarch/cpu.c b/libavutil/loongarch/cpu.c i

Re: [FFmpeg-devel] [PATCH 3/4] avcodec/mobiclip: Check input size before (re)allocation

Please ignore fail info form loongarch64 patchwork temporarily. I have stoped the service from Nov 16th, but there are still result post to patchwork. Have mailed to Andriy to help analyze this problem together. > 2022年11月19日 05:09，Michael Niedermayer 写道： > > Fixes: Timeout > Fixes: > 52566/cl

Re: [FFmpeg-devel] [PATCH v1 3/3] swscale/la: Add output_lasx.c file.

> 2022年9月11日 10:06，Shiyou Yin 写道： > > > >> 2022年9月9日 21:11，Andreas Rheinhardt > <mailto:andreas.rheinha...@outlook.com>> 写道： >> >> Shiyou Yin: >>> >>> >>>> 2022年9月6日 16:12，Shiyou Yin >>> <mailto:y

Re: [FFmpeg-devel] [PATCH v1 3/3] swscale/la: Add output_lasx.c file.

> 2022年9月9日 21:11，Andreas Rheinhardt 写道： > > Shiyou Yin: >> >> >>> 2022年9月6日 16:12，Shiyou Yin 写道： >>> >>>> >>>> 2022年8月29日 20:30，Andreas Rheinhardt >>> <mailto:andreas.rheinha...@outlook.com>> 写道： >>

Re: [FFmpeg-devel] Fix bugs on Mips platform.

> 2022年9月9日 17:41，Hao Chen 写道： > > v2: Modifies the format of the Commit message. > > [PATCH v2 1/2] lavc/mips: Fix bugs in me_cmp_msa.c file. > [PATCH v2 2/2] lavc/mips: Fix hevc decoding bugs on MIPS paltform. > > LGTM. ___ ffmpeg-devel mailing

Re: [FFmpeg-devel] [PATCH v1 3/3] swscale/la: Add output_lasx.c file.

> 2022年9月6日 16:12，Shiyou Yin 写道： > >> >> 2022年8月29日 20:30，Andreas Rheinhardt > <mailto:andreas.rheinha...@outlook.com>> 写道： >> >> Hao Chen: >>> ffmpeg -i ~/media/1_h264_1080p_30fps_3Mbps.mp4 -f rawvideo -s 640x480 >>> -pix_fmt >

Re: [FFmpeg-devel] [PATCH v2] avcodec/mips: Fix MMI macro replaces in HEVC Decoder

> 2022年8月18日 20:29，Shiyou Yin 写道： > > > >> 2022年8月18日 19:44，戚铁铮 mailto:qitiezh...@360.cn>> 写道： >> >> >> At 2022/8/18 PM 7:01, "Qi Tiezheng" > <mailto:qitiezh...@360.cn>> wrote: >> >>> The latest commit of Lo

Re: [FFmpeg-devel] [PATCH v1 3/3] swscale/la: Add output_lasx.c file.

> 2022年8月29日 20:30，Andreas Rheinhardt 写道： > > Hao Chen: >> ffmpeg -i ~/media/1_h264_1080p_30fps_3Mbps.mp4 -f rawvideo -s 640x480 >> -pix_fmt >> rgb24 -y /dev/null -an >> before: 150fps >> after: 183fps >> >> Signed-off-by: Hao Chen >> --- >> libswscale/loongarch/Makefile |

Re: [FFmpeg-devel] [PATCH v1 2/3] swscale/la: Add yuv2rgb_lasx.c and rgb2rgb_lasx.c files

> --- /dev/null > +++ b/libswscale/loongarch/rgb2rgb_lasx.c > @@ -0,0 +1,52 @@ > +/* > + * Copyright (c) 2022 Loongson Technology Corporation Limited > + * Contributed by Hao Chen(chen...@loongson.cn) > + * > + * This file is part of FFmpeg. > + * > + * FFmpeg is free software; you can redistribu

Re: [FFmpeg-devel] [PATCH v1 1/3] swscale/la: Optimize hscale functions with lasx.

> + > +void planar_rgb_to_uv_lasx(uint8_t *_dstU, uint8_t *_dstV, const uint8_t > *src[4], > + int width, int32_t *rgb2yuv) > +{ > +int i; > +uint16_t *dstU = (uint16_t *)_dstU; > +uint16_t *dstV = (uint16_t *)_dstV; > +int set = 0x4001 << (R

Re: [FFmpeg-devel] [PATCH v2] avcodec/mips: Fix MMI macro replaces in HEVC Decoder

> 2022年8月18日 19:44，戚铁铮 写道： > > > At 2022/8/18 PM 7:01, "Qi Tiezheng" > wrote: > >> The latest commit of Loongson MMI macro replaces were incorrect. >> It makes a mass of green tints on HEVC videos when playing. I've >> compared it with the older MMI implementation,

[FFmpeg-devel] [PATCH] MAINTAINERS: add myself as maintainer for LoongArch.

Manojkumar Bhosale, Shiyou Yin +LoongArch Shiyou Yin Mac OS X / PowerPC Romain Dolbeau, Guillaume Poirier Amiga / PowerPC Colin Ward Linux / PowerPC Lauri Kasanen -- 2.20.1

Re: [FFmpeg-devel] Optimize VP8,VP9,WMV3 decoding for loongarch.

> 2021年12月18日 下午10:27，Hao Chen 写道： > > ./ffmpeg -i ../9_vp8_1080p_30fps_2Mbps.webm -f rawvideo -y /dev/null -an > before: 210fps > after : 585fps > ffmpeg -i ../10_vp9_1080p_30fps_3Mbps.webm -f rawvideo -y /dev/null -an > before:170fps > after :294fps > ./ffmpeg -i 11_wmv3_720p_24fps_7Mbps.wmv

Re: [FFmpeg-devel] [PATCH v2 2/7] avcodec: [loongarch] Optimize h264_chroma_mc with LASX.

> 2021年12月15日 上午12:26，Michael Niedermayer 写道： > > On Tue, Dec 14, 2021 at 09:33:11PM +0800, Hao Chen wrote: >> From: Shiyou Yin >> >> ./ffmpeg -i ../1_h264_1080p_30fps_3Mbps.mp4 -f rawvideo -y /dev/null -an >> before:1

Re: [FFmpeg-devel] Optimize H264 decoding for loongarch.

> 2021年12月15日 上午11:51，Hao Chen 写道： > > ./ffmpeg -i 1_h264_1080p_30fps_3Mbps.mp4 -f rawvideo -y /dev/null -an > before: 170fps > after : 296fps > > V2: Update loongson_intrinsics.h to 1.0.3 in patch 2/7. > V3: Resubmit these patches due to the encoding problem of the last > email. > > [PATCH v

Re: [FFmpeg-devel] Optimize H264 decoding for loongarch.

> 2021年12月14日 下午9:33，Hao Chen 写道： > > ./ffmpeg -i 1_h264_1080p_30fps_3Mbps.mp4 -f rawvideo -y /dev/null -an > before: 170fps > after : 296fps > > V2: Update loongson_intrinsics.h to 1.0.3 in patch 2/7. > > [PATCH v2 1/7] avutil: [loongarch] Add support for loongarch SIMD. > [PATCH v2 2/7] avc

[FFmpeg-devel] [PATCH v2 3/3] avcodec: [loongarch] Optimize decode_significance/_8x8_loongarch.

From: Hao Chen Decoding 1080P H264 from 168fps to 170fps. Signed-off-by: Shiyou Yin --- libavcodec/h264_cabac.c | 2 + libavcodec/loongarch/h264_cabac.c | 140 ++ 2 files changed, 142 insertions(+) create mode 100644 libavcodec/loongarch/h264_cabac.c

[FFmpeg-devel] [PATCH v2 1/3] configure: Add support for loongarch.

For la464 cpu: ./configure --cpu=la464 With cross-compiler: ./configure --cross-prefix=loongarch64-linux-gnu- \ --enable-cross-compile --arch=loongarch64 \ --target-os=linux --cpu=la464 --- Changelog | 1 + configure | 23 +++ 2 files changed, 24 inser

[FFmpeg-devel] Add support for loongarch.

V2: 1. rebase. 2. Change Author email from yinshi...@loongson.cn to yinshiyou...@loongson.cn for 1/3 2/3. 3. Refine 2/3. [PATCH v2 1/3] configure: Add support for loongarch. [PATCH v2 2/3] avcodec: [loongarch] optimize get_cabac. [PATCH v2 3/3] avcodec: [loongarch] Optimize decode_significance. _

[FFmpeg-devel] [PATCH v2 2/3] avcodec: [loongarch] optimize get_cabac.

b/libavcodec/loongarch/cabac.h new file mode 100644 index 00..e1c946fe16 --- /dev/null +++ b/libavcodec/loongarch/cabac.h @@ -0,0 +1,238 @@ +/* + * Loongson optimized cabac + * + * Copyright (c) 2020 Loongson Technology Corporation Limited + * Contributed by Shiyou Yin + *

[FFmpeg-devel] [PATCH 3/3] avcodec: [loongarch] Optimize decode_significance/_8x8_loongarch.

From: Hao Chen Decoding 1080P H264 from 168fps to 170fps. Signed-off-by: Shiyou Yin --- libavcodec/h264_cabac.c | 2 + libavcodec/loongarch/h264_cabac.c | 140 ++ 2 files changed, 142 insertions(+) create mode 100644 libavcodec/loongarch/h264_cabac.c

[FFmpeg-devel] [PATCH 2/3] avcodec: [loongarch] optimize get_cabac.

b/libavcodec/loongarch/cabac.h new file mode 100644 index 00..71e8ba3be7 --- /dev/null +++ b/libavcodec/loongarch/cabac.h @@ -0,0 +1,254 @@ +/* + * Loongson optimized cabac + * + * Copyright (c) 2020 Loongson Technology Corporation Limited + * Contributed by Shiyou Yin + *

[FFmpeg-devel] [PATCH 1/3] configure: Add support for loongarch.

For la464 cpu: ./configure --cpu=la464 With cross-compiler: ./configure --cross-prefix=loongarch64-linux-gnu- \ --enable-cross-compile --arch=loongarch64 \ --target-os=linux --cpu=la464 --- Changelog | 1 + configure | 17 + 2 files changed, 18 insertions(+) diff --git a/Changelo

[FFmpeg-devel] Add support for loongarch.

[PATCH 1/3] configure: Add support for loongarch. [PATCH 2/3] avcodec: [loongarch] optimize get_cabac. [PATCH 3/3] avcodec: [loongarch] Optimize decode_significance. ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinf

[FFmpeg-devel] [PATCH v4 5/5] mips: Fix potential illegal instruction error.

MSA2 optimizations are attached to MSA macros in generic_macros_msa.h. It's difficult to do runtime check for them. Remove this part of code can make it more robust. H264 1080p decoding: 5.13x==>5.12x. --- configure | 7 +-- libavutil/mips/generic_macros_msa.h | 37 -

[FFmpeg-devel] [PATCH v4 3/5] avcodec/mips: Optimize function ff_h264_loop_filter_strength_msa.

From: gxw Speed of decoding H264 1080P: 5.05x ==> 5.13x Signed-off-by: Shiyou Yin --- libavcodec/mips/Makefile| 3 +- libavcodec/mips/h264_deblock_msa.c | 153 libavcodec/mips/h264dsp_init_mips.c | 2 + libavcodec/mips/h264dsp_mip

[FFmpeg-devel] [PATCH v4 4/5] avcodec/mips: Refine ff_h264_h_lpf_luma_inter_msa

From: gxw Using mask to avoid judgment, H264 4K decoding speed improved about 0.1fps tested on 3A4000 Signed-off-by: Shiyou Yin --- libavcodec/mips/h264dsp_msa.c | 465 -- 1 file changed, 171 insertions(+), 294 deletions(-) diff --git a/libavcodec/mips

[FFmpeg-devel] [PATCH V4] [mips] Optimize H264 decoding for MIPS platform.

v2: Fixed a build error in [PATCH 2/5]. v3: add patch 4/5. v4: Fix bug in 2/5 caused by instruction 'lhu' on BIGENDIAN environment. [PATCH v4 1/5] avcodec/mips: Restore the initialization sequence of [PATCH v4 2/5] avcodec/mips: Refine get_cabac_inline_mips. [PATCH v4 3/5] avcodec/mips: Optimize f

[FFmpeg-devel] [PATCH v4 1/5] avcodec/mips: Restore the initialization sequence of MSA and MMI in ff_h264chroma_init_mips.

The MSA optimization has been refined in commit 93218c2 and ce0a52e. It is better than MMI version now. Speed of decoding H264: 4.83x ==> 4.89x (tested on 3A4000). --- libavcodec/mips/h264chroma_init_mips.c | 19 +-- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/li

[FFmpeg-devel] [PATCH v4 2/5] avcodec/mips: Refine get_cabac_inline_mips.

(c) 2018 Shiyou Yin + * Contributed by Shiyou Yin + *Gu Xiwei(guxiwei...@loongson.cn) * * This file is part of FFmpeg. * @@ -25,18 +26,18 @@ #define AVCODEC_MIPS_CABAC_H #include "libavcodec/cabac.h" -#include "libavutil/mips/asmdefs.h" +#include &qu

[FFmpeg-devel] [PATCH v3 5/5] mips: Fix potential illegal instruction error.

MSA2 optimizations are attached to MSA macros in generic_macros_msa.h. It's difficult to do runtime check for them. Remove this part of code can make it more robust. H264 1080p decoding: 5.13x==>5.12x. --- configure | 7 +-- libavutil/mips/generic_macros_msa.h | 37 -

[FFmpeg-devel] [PATCH v3 4/5] avcodec/mips: Refine ff_h264_h_lpf_luma_inter_msa

From: gxw Using mask to avoid judgment, H264 4K decoding speed improved about 0.1fps tested on 3A4000 Signed-off-by: Shiyou Yin --- libavcodec/mips/h264dsp_msa.c | 465 -- 1 file changed, 171 insertions(+), 294 deletions(-) diff --git a/libavcodec/mips

[FFmpeg-devel] [PATCH v3 3/5] avcodec/mips: Optimize function ff_h264_loop_filter_strength_msa.

From: gxw Speed of decoding H264 1080P: 5.05x ==> 5.13x Signed-off-by: Shiyou Yin --- libavcodec/mips/Makefile| 3 +- libavcodec/mips/h264_deblock_msa.c | 153 libavcodec/mips/h264dsp_init_mips.c | 2 + libavcodec/mips/h264dsp_mip

[FFmpeg-devel] [PATCH v3 2/5] avcodec/mips: Refine get_cabac_inline_mips.

(c) 2018 Shiyou Yin + * Contributed by Shiyou Yin + *Gu Xiwei(guxiwei...@loongson.cn) * * This file is part of FFmpeg. * @@ -25,18 +26,18 @@ #define AVCODEC_MIPS_CABAC_H #include "libavcodec/cabac.h" -#include "libavutil/mips/asmdefs.h" +#include &qu

[FFmpeg-devel] [mips] Optimize H264 decoding for MIPS platform.

v2: Fixed a build error in [PATCH 2/5]. v3: add patch 4/5. [PATCH v3 1/5] avcodec/mips: Restore the initialization sequence of [PATCH v3 2/5] avcodec/mips: Refine get_cabac_inline_mips. [PATCH v3 3/5] avcodec/mips: Optimize function ff_h264_loop_filter_strength_msa. [PATCH v3 4/5] avcodec/mips: Re

[FFmpeg-devel] [PATCH v3 1/5] avcodec/mips: Restore the initialization sequence of MSA and MMI in ff_h264chroma_init_mips.

The MSA optimization has been refined in commit 93218c2 and ce0a52e. It is better than MMI version now. Speed of decoding H264: 4.83x ==> 4.89x (tested on 3A4000). --- libavcodec/mips/h264chroma_init_mips.c | 19 +-- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/li

[FFmpeg-devel] [PATCH v3 3/4] avcodec/mips: Optimize function ff_h264_loop_filter_strength_msa.

From: gxw Speed of decoding H264: 5.45x ==> 5.53x Signed-off-by: Shiyou Yin --- libavcodec/mips/Makefile| 3 +- libavcodec/mips/h264_deblock_msa.c | 153 libavcodec/mips/h264dsp_init_mips.c | 2 + libavcodec/mips/h264dsp_mips.h |

[FFmpeg-devel] [PATCH v3 4/4] Fix potential illegal instruction error.

MSA2 optimizations are attached to MSA macros in generic_macros_msa.h. It's difficult to do runtime check for them. Remove this part of code can make it more robust. Impact on performance is not obvious(167fps==>166fps). --- configure | 7 +-- libavutil/mips/generic_

[FFmpeg-devel] MIPS: Optimize H264 decoding.

H264 decoding speed: 154fps ==> 165fps, 5.14x ==> 5.53x (tested on 3A4000) V2: Fixed a build error in [PATCH 2/3]. "Error: opcode not supported on this processor: mips32r2 (mips32r2) `dsbh $10,$10'" V3: Add a fix patch to make MSA optimization more robust. [PATCH v3 1/4] avcodec/mips: Restore t

[FFmpeg-devel] [PATCH v3 1/4] avcodec/mips: Restore the initialization sequence of MSA and MMI in ff_h264chroma_init_mips.

The MSA version has been refined in commit 93218c2 and ce0a52e, and is better than MMI version now. Speed of decoding H264: 5.14x ==> 5.23x (tested on 3A4000). --- libavcodec/mips/h264chroma_init_mips.c | 19 +-- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/libavc

[FFmpeg-devel] [PATCH v3 2/4] avcodec/mips: Refine get_cabac_inline_mips.

(c) 2018 Shiyou Yin + * Contributed by Shiyou Yin + *Gu Xiwei(guxiwei...@loongson.cn) * * This file is part of FFmpeg. * @@ -25,18 +26,18 @@ #define AVCODEC_MIPS_CABAC_H #include "libavcodec/cabac.h" -#include "libavutil/mips/asmdefs.h" +#include &qu

Re: [FFmpeg-devel] [PATCH 1/2] avcodec/mips: [loongson] Fixed mmi optimization

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Shiyou Yin >Sent: Thursday, September 3, 2020 2:30 PM >To: ffmpeg-devel@ffmpeg.org >Cc: gxw >Subject: [FFmpeg-devel] [PATCH 1/2] avcodec/mi

Re: [FFmpeg-devel] [PATCH 2/2] Fix msa can't be disabled when '--cpu=loongson3a' assigned.

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Shiyou Yin >Sent: Thursday, September 3, 2020 2:30 PM >To: ffmpeg-devel@ffmpeg.org >Subject: [FFmpeg-devel] [PATCH 2/2] Fix msa can't be disable

[FFmpeg-devel] [PATCH v2 3/3] avcodec/mips: Optimize function ff_h264_loop_filter_strength_msa.

From: gxw Speed of decoding H264: 5.45x ==> 5.53x Signed-off-by: Shiyou Yin --- libavcodec/mips/Makefile| 3 +- libavcodec/mips/h264_deblock_msa.c | 153 libavcodec/mips/h264dsp_init_mips.c | 2 + libavcodec/mips/h264dsp_mips.h |

[FFmpeg-devel] [PATCH v2 1/3] avcodec/mips: Restore the initialization sequence of MSA and MMI in ff_h264chroma_init_mips.

Speed of decoding H264: 5.14x ==> 5.23x (tested on 3A4000). --- libavcodec/mips/h264chroma_init_mips.c | 19 +-- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/libavcodec/mips/h264chroma_init_mips.c b/libavcodec/mips/h264chroma_init_mips.c index 6bb19d3..755cc04 10

[FFmpeg-devel] [PATCH v2 2/3] avcodec/mips: Refine get_cabac_inline_mips.

(c) 2018 Shiyou Yin + * Contributed by Shiyou Yin + *Gu Xiwei(guxiwei...@loongson.cn) * * This file is part of FFmpeg. * @@ -25,18 +26,18 @@ #define AVCODEC_MIPS_CABAC_H #include "libavcodec/cabac.h" -#include "libavutil/mips/asmdefs.h" +#include &qu

[FFmpeg-devel] [loongson] Optimize H264 decoding.

H264 decoding speed: 154fps ==> 165fps, 5.14x ==> 5.53x (tested on 3A4000) V2: Fixed a build error in [PATCH 2/3]. "Error: opcode not supported on this processor: mips32r2 (mips32r2) `dsbh $10,$10'" [PATCH 1/3] avcodec/mips: Restore the initialization sequence of MSA. [PATCH 2/3] avcodec/mips:

[FFmpeg-devel] [PATCH 3/3] avcodec/mips: Optimize function ff_h264_loop_filter_strength_msa.

From: gxw Speed of decoding H264: 5.45x ==> 5.53x Signed-off-by: Shiyou Yin --- libavcodec/mips/Makefile| 3 +- libavcodec/mips/h264_deblock_msa.c | 153 libavcodec/mips/h264dsp_init_mips.c | 2 + libavcodec/mips/h264dsp_mips.h |

[FFmpeg-devel] [loongson] Optimize H264 decoding.

H264 decoding speed: 154fps ==> 165fps, 5.14x ==> 5.53x (tested on 3A4000) [PATCH 1/3] avcodec/mips: Restore the initialization sequence of MSA. [PATCH 2/3] avcodec/mips: Refine get_cabac_inline_mips. [PATCH 3/3] avcodec/mips: Optimize function ff_h264_loop_filter_strength_msa. _

[FFmpeg-devel] [PATCH 1/3] avcodec/mips: Restore the initialization sequence of MSA and MMI in ff_h264chroma_init_mips.

Speed of decoding H264: 5.14x ==> 5.23x (tested on 3A4000). --- libavcodec/mips/h264chroma_init_mips.c | 19 +-- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/libavcodec/mips/h264chroma_init_mips.c b/libavcodec/mips/h264chroma_init_mips.c index 6bb19d3..755cc04 10

[FFmpeg-devel] [PATCH 2/3] avcodec/mips: Refine get_cabac_inline_mips.

(c) 2018 Shiyou Yin + * Contributed by Shiyou Yin + *Gu Xiwei(guxiwei...@loongson.cn) * * This file is part of FFmpeg. * @@ -25,18 +26,18 @@ #define AVCODEC_MIPS_CABAC_H #include "libavcodec/cabac.h" -#include "libavutil/mips/asmdefs.h" +#include &qu

[FFmpeg-devel] [PATCH 2/2] Fix msa can't be disabled when '--cpu=loongson3a' assigned.

There are compiler and runtime check for MSA and MMI. Remove the redundant setting of MSA and MMI for cores specified by "--cpu". Signed-off-by: Shiyou Yin --- configure | 9 - 1 file changed, 9 deletions(-) diff --git a/configure b/configure index 5640720..7f103fa 10

[FFmpeg-devel] [PATCH 1/2] avcodec/mips: [loongson] Fixed mmi optimization

From: gxw Test case fate-checkasm-h264pred failed in latest community code. This patch fixed the bug. Signed-off-by: Shiyou Yin --- libavcodec/mips/h264pred_mmi.c | 8 ++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/libavcodec/mips/h264pred_mmi.c b/libavcodec/mips

[FFmpeg-devel] [PATCH 4/4] [mips]: Fix segfault in imdct36_mips_float.

'li.s' is a synthesized instruction, it does not work properly when compiled with clang on mips, and A segfault occurred. --- libavcodec/mips/aacpsdsp_mips.c | 13 +- libavcodec/mips/aacpsy_mips.h | 14 +- libavcodec/mips/fft_mips.c| 12 +- libavcodec/mips/

[FFmpeg-devel] [PATCH 3/4] [mips]: Fix a bug in get_cabac_inline_mips.

Failed fate case: fate-h264-conformance-caba2_sony_e Clang is more strict in the use of register constraint. --- libavcodec/mips/cabac.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/libavcodec/mips/cabac.h b/libavcodec/mips/cabac.h index c595915..3d09e93 100644 --- a/libavco

[FFmpeg-devel] [PATCH 0/4] [mips]: Adapt clang compiler for mips.

Fixed four prob encountered when compiling ffmpeg with clang on mips. configure: ./configure --disable-mmi --cc=clang --cxx=clang++ --ld=clang [PATCH 1/4] [mips]: Fix register constraint error reported by clang. [PATCH 2/4] [mips]: Fix prob that 'ulw' and 'uld' unsupported by clang. [PATCH 3/4] [

[FFmpeg-devel] [PATCH 1/4] [mips]: Fix register constraint error reported by clang.

Clang report following error in aacsbr_mips.c,ac3dsp_mips.c and aacdec_mips.c: "couldn't allocate output register for constraint 'r'" Use 'f' constraint for float variable. --- libavcodec/mips/aacdec_mips.c | 2 +- libavcodec/mips/aacsbr_mips.c | 2 +- libavcodec/mips/sbrdsp_mips.c | 19 +++

[FFmpeg-devel] [PATCH 2/4] [mips]: Fix prob that 'ulw' and 'uld' unsupported by clang.

GCC support these two synthesized instruction, but clang does not yet. Use machine instruction instead to adapt clang compiler. --- libavutil/mips/generic_macros_msa.h | 14 -- 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/libavutil/mips/generic_macros_msa.h b/libavuti

Re: [FFmpeg-devel] [PATCH v6 0/6] MIPS MSA & MMI Runtime detection support

IPS MSA & MMI Runtime detection >support > >On Thu, Jul 23, 2020 at 04:18:32PM +0200, Michael Niedermayer wrote: >> On Tue, Jul 21, 2020 at 03:40:25PM +0800, Shiyou Yin wrote: >> > >-Original Message- >> > >From: ffmpeg-devel-boun...@ffmpeg.org &g

Re: [FFmpeg-devel] [PATCH v6 0/6] MIPS MSA & MMI Runtime detection support

& MMI Runtime detection support > >This series adds MIPS MSA & MMI runtime detection support > >Please review. > >Thanks! > >v2: > - Add CPUCFG support. > - Add "-mloongson-ext" to MMIFLAGS for Loongson-3 as well. (Loongson2F don't > need

[FFmpeg-devel] [PATCH] avcodec/mips: fix type mismatch in h264dsp_msa.c

gcc warning: assignment from incompatible pointer type. --- libavcodec/mips/h264dsp_mips.h | 24 +-- libavcodec/mips/h264dsp_msa.c | 94 ++ 2 files changed, 62 insertions(+), 56 deletions(-) diff --git a/libavcodec/mips/h264dsp_mips.h b/libavcodec/

Re: [FFmpeg-devel] [PATCH v5 5/6] libavcodec: MIPS: MMI: Fix type mismatches

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Thursday, July 2, 2020 11:46 PM >To: ffmpeg-devel@ffmpeg.org >Cc: Jiaxun Yang >Subject: [FFmpeg-devel] [PATCH v5 5/6] libavcodec: MIPS: MMI: Fix type >mism

Re: [FFmpeg-devel] [PATCH v5 4/6] libavcodec: Enable runtime detection for MIPS MMI & MSA

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Thursday, July 2, 2020 11:46 PM >To: ffmpeg-devel@ffmpeg.org >Cc: Jiaxun Yang >Subject: [FFmpeg-devel] [PATCH v5 4/6] libavcodec: Enable runtime detection

Re: [FFmpeg-devel] [PATCH v5 1/6] ffbuild: Refine MIPS handling

Attachment patch include all changes corresponding to my following comments, for your reference. >-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Thursday, July 2, 2020 11:46 PM >To: ffmpeg-devel@ffmpeg.

Re: [FFmpeg-devel] [PATCH v4 4/4] libavcodec: Enable runtime detection for MIPS MMI & MSA

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Monday, June 8, 2020 11:32 AM >To: ffmpeg-devel@ffmpeg.org >Cc: yinshi...@loongson.cn; Jiaxun Yang >Subject: [FFmpeg-devel] [PATCH v4 4/4] libavcodec: Enabl

Re: [FFmpeg-devel] [PATCH v4 3/4] libavutil: Detect MMI and MSA flags for MIPS

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Monday, June 8, 2020 11:32 AM >To: ffmpeg-devel@ffmpeg.org >Cc: yinshi...@loongson.cn; Jiaxun Yang >Subject: [FFmpeg-devel] [PATCH v4 3/4] libavutil: Detect

Re: [FFmpeg-devel] [PATCH v4 2/4] libavutils: Add parse_r helper for MIPS

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >jiaxun.y...@flygoat.com >Sent: Tuesday, June 9, 2020 1:59 PM >To: Shiyou Yin; 'FFmpeg development discussions and patches' >Subject: R

Re: [FFmpeg-devel] [PATCH v4 2/4] libavutils: Add parse_r helper for MIPS

>-Original Message- >From: jiaxun.y...@flygoat.com [mailto:jiaxun.y...@flygoat.com] >Sent: Tuesday, June 9, 2020 2:03 AM >To: FFmpeg development discussions and patches; Shiyou Yin; 'FFmpeg >development discussions and >patches' >Subject: Re: [FFmpeg-devel]

Re: [FFmpeg-devel] [PATCH v4 2/4] libavutils: Add parse_r helper for MIPS

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Monday, June 8, 2020 11:30 AM >To: ffmpeg-devel@ffmpeg.org >Cc: yinshi...@loongson.cn; Jiaxun Yang >Subject: [FFmpeg-devel] [PATCH v4 2/4] libavutils: Add p

Re: [FFmpeg-devel] [PATCH v4 1/4] ffbuild: Refine MIPS handling

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Monday, June 8, 2020 11:30 AM >To: ffmpeg-devel@ffmpeg.org >Cc: yinshi...@loongson.cn; Jiaxun Yang >Subject: [FFmpeg-devel] [PATCH v4 1/4] ffbuild: Refine M

Re: [FFmpeg-devel] [PATCH v3 1/4] ffbuild: Refine MIPS handling

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Saturday, June 6, 2020 3:34 PM >To: ffmpeg-devel@ffmpeg.org >Cc: yinshi...@loongson.cn; Jiaxun Yang >Subject: [FFmpeg-devel] [PATCH v3 1/4] ffbuild: Refine

Re: [FFmpeg-devel] [PATCH v2 1/4] ffbuild: Refine MIPS handling

For your convenience, I add the previous comments into this patch. >-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Tuesday, June 2, 2020 10:15 PM >To: ffmpeg-devel@ffmpeg.org >Cc: yinshi...@loongson.cn;

Re: [FFmpeg-devel] [PATCH 1/3] ffbuild: Refine MIPS handling

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Shiyou Yin >Sent: Sunday, May 31, 2020 11:33 AM >To: 'FFmpeg development discussions and patches' >Cc: yinshi...@loongson.cn >Subject: Re:

Re: [FFmpeg-devel] [PATCH 1/3] ffbuild: Refine MIPS handling

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Tuesday, May 26, 2020 5:48 PM >To: ffmpeg-devel@ffmpeg.org >Cc: yinshi...@loongson.cn; Jiaxun Yang >Subject: [FFmpeg-devel] [PATCH 1/3] ffbuild: Refine MIPS

Re: [FFmpeg-devel] [PATCH 1/3] ffbuild: Refine MIPS handling

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Jiaxun Yang >Sent: Tuesday, May 26, 2020 5:48 PM >To: ffmpeg-devel@ffmpeg.org >Cc: yinshi...@loongson.cn; Jiaxun Yang >Subject: [FFmpeg-devel] [PATCH 1/3] ffbuild: Refine MI

Re: [FFmpeg-devel] [PATCH v2] avcodec/mips: msa optimizations for vc1dsp

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of gxw >Sent: Monday, October 21, 2019 3:57 PM >To: ffmpeg-devel@ffmpeg.org >Subject: [FFmpeg-devel] [PATCH v2] avcodec/mips: msa optimizations for vc1dsp > >Performance of WMV3 d

Re: [FFmpeg-devel] [PATCH] avcodec/mips: msa optimizations for vc1dsp

>diff --git a/libavcodec/mips/vc1dsp_msa.c b/libavcodec/mips/vc1dsp_msa.c >new file mode 100644 >index 000..1619ea4 >--- /dev/null >+++ b/libavcodec/mips/vc1dsp_msa.c >@@ -0,0 +1,483 @@ >+/* >+ * Loongson SIMD optimized vc1dsp >+ * >+ * Copyright (c) 2019 Loongson Technology Corporation Limited

Re: [FFmpeg-devel] [PATCH] avcodec/mips: Fixed four warnings in vc1dsp

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of gxw >Sent: Saturday, October 12, 2019 10:48 AM >To: ffmpeg-devel@ffmpeg.org >Subject: [FFmpeg-devel] [PATCH] avcodec/mips: Fixed four warnings in vc1dsp > >Change the stride ar

[FFmpeg-devel] [PATCH] avcodec/mips: Fix a warnning of indentation not reflect the block structure.

The indentation of code dose not reflect the if block structure in 'apply_ltp_mips', and this will generate a warnning when build with '-Wall' or '-Wmisleading-indentation'. --- libavcodec/mips/aacdec_mips.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/libavcodec/mips/

[FFmpeg-devel] [PATCH] avutil/mips: remove redundant code in TRANSPOSE16x8_UB_UB.

--- libavutil/mips/generic_macros_msa.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/libavutil/mips/generic_macros_msa.h b/libavutil/mips/generic_macros_msa.h index 9ac0583..219ff07 100644 --- a/libavutil/mips/generic_macros_msa.h +++ b/libavutil/mips/generic_macros_msa.h @@ -2523,8 +2523

Re: [FFmpeg-devel] [PATCH v4] avutil/mips: refine msa macros CLIP_*.

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of >Michael Niedermayer >Sent: Friday, August 9, 2019 12:07 AM >To: FFmpeg development discussions and patches >Subject: Re: [FFmpeg-devel] [PATCH v4] avutil/mips: refine msa macr

Re: [FFmpeg-devel] [PATCH v4] avutil/mips: refine msa macros CLIP_*.

LGTM. >-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of gxw >Sent: Wednesday, August 7, 2019 5:52 PM >To: ffmpeg-devel@ffmpeg.org >Subject: [FFmpeg-devel] [PATCH v4] avutil/mips: refine msa macros CLIP_*. > >Changing details

Re: [FFmpeg-devel] [PATCH v2] avutil/mips: refactor msa SLDI_Bn_0 and SLDI_Bn macros.

LGTM. >-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of gxw >Sent: Tuesday, August 6, 2019 7:11 PM >To: ffmpeg-devel@ffmpeg.org >Subject: [FFmpeg-devel] [PATCH v2] avutil/mips: refactor msa SLDI_Bn_0 and >SLDI_Bn macros. > >

Re: [FFmpeg-devel] [PATCH] avutil/mips: refactor msa SLDI_Bn_0 and SLDI_Bn macros.

>-Original Message- >From: ffmpeg-devel-boun...@ffmpeg.org [mailto:ffmpeg-devel-boun...@ffmpeg.org] >On Behalf Of gxw >Sent: Tuesday, August 6, 2019 11:38 AM >To: ffmpeg-devel@ffmpeg.org >Subject: [FFmpeg-devel] [PATCH] avutil/mips: refactor msa SLDI_Bn_0 and >SLDI_Bn macros. > >Changing

  1   2   >